Wittgenstein: strong correlationist?

Last week I was following the debates on Wittgenstein's middle period triggered by a forthcoming book by Mauro Engelmann (Wittgenstein's Philosophical Development). I couldn't avoid having in mind Meillassoux charge of strong correlationism addressed to Wittgenstein (and Heidegger). It seems like Wittgenstein, in the middle of a very interesting sequence of philosophical moves, was progressively recoiling towards a philosophy confined to correlations. It is however unclear whether his is a case of strong correlationism or rather one of what Meillassoux calls metaphysics of subjectivity. Indeed, few years back I lectured a course on Hegel and Wittgenstein where we examined the Investigations together with the Phenomenology under the light of not only Brandom's work (it was before the Spirit of Trust) and McDowell's hints but also the pioneer book by David Lamb from 1980 (Language and Perception in Hegel and Wittgenstein). At the time I was convinced that the similarities between the two approaches to the transcendental distinction (which is crucial in the contrast between correlationisms and metaphysics of subjectivity as I see it) were relevantly similar. But surely there are differences, the Investigations seems indeed restricted to the way we go, to our practices, as thought is not conceivable outside them. When he insists on the unintelligibility of something beyond our practices he was not thinking about a general structure of thought that could be appreciated even though we are locked in our practices. Even in the Remarks on the Foundation of Mathematics, when he talks about God having to do mathematics in order to determine something mathematical, he seems to be pointing at mathematics as having no results that could be intelligible to our practices without appeal to those practices and not at something defined by our practices.

Still, it maybe makes sense that in the middle period (when he wrote the Remarks) he got eventually closer to a metaphysics of subjectivity. It is interesting that first he abandoned the formalism of the Tractatus (because of the issue with logical spaces) in favour of a phenomenological formalism that could still have room for a distinction between our phenomenology and the grammar of the formalism. He gradually adopted a thoroughly anthropological take that has no space for an independent formalism of any sort. Then the genetic method becomes more important than grammar (in the Brown Book, around 1934). With no formalism - and no importance attached to grammar - we are left with the idea that our ways are all we can reach in our thoughts (and thinking practices). If they are facts, those are entirely within the correlation. Such a view can be found in the Investigations. It is not the only one, but if we stress it, the book really sound like strong correlationism.